Wine Quality Predictionand Deployment

Statsly Analytics

By Moumita Das

Problem Statement

The red wine industry shows a recent exponential growth as social drinking is on the rise. Nowadays, industryplayers are using product quality certifications to promote their products. This is a time-consuming processand requires the assessment given by human experts, which makes this process very expensive. Also, the price of red wine depends on a rather abstract concept of wine appreciation by wine tasters,opinions among whom may have a high degree of variability. Another vital factor in red wine certification andquality assessment is physicochemical tests, which are laboratory-based and consider factors like acidity, pHlevel, sugar, and other chemical properties. The red wine market would be of interest if the human quality oftasting can be related to wine’s chemical properties so that certification and quality assessment andassurance processes are more controlled. Determine which features are the best quality red wine indicators and generate insights into each of thesefactors to our model’s red wine quality. Predict the quality of wine on the basis of giving features. Deploy themodel.

Importing all libraries

import pandas as pd # for reading the file
import numpy as np # for numericals & mathematicals experssion
import seaborn as sns # for data visulization
import matplotlib.pyplot as plt # for data visulization
%matplotlib inline 
import time 
import warnings
warnings.filterwarnings('ignore')
from sklearn import metrics
import seaborn as sns

Import Dataset

data = pd.read_csv("Wine_Quality_Predictionand_Deployment_TASK_6.csv")

Top 5 rows of the Dataset

 data.head()

	fixed acidity	volatile acidity	citric acid	residual sugar	chlorides	free sulfur dioxide	total sulfur dioxide	density	pH	sulphates	alcohol	quality
0	7.4	0.70	0.00	1.9	0.076	11.0	34.0	0.9978	3.51	0.56	9.4	5
1	7.8	0.88	0.00	2.6	0.098	25.0	67.0	0.9968	3.20	0.68	9.8	5
2	7.8	0.76	0.04	2.3	0.092	15.0	54.0	0.9970	3.26	0.65	9.8	5
3	11.2	0.28	0.56	1.9	0.075	17.0	60.0	0.9980	3.16	0.58	9.8	6
4	7.4	0.70	0.00	1.9	0.076	11.0	34.0	0.9978	3.51	0.56	9.4	5

pip install pandas-profiling

Requirement already satisfied: pandas-profiling in c:\users\dell\anaconda3\lib\site-packages (3.1.0)
Requirement already satisfied: matplotlib>=3.2.0 in c:\users\dell\anaconda3\lib\site-packages (from pandas-profiling) (3.4.3)
Requirement already satisfied: requests>=2.24.0 in c:\users\dell\anaconda3\lib\site-packages (from pandas-profiling) (2.26.0)
Requirement already satisfied: joblib~=1.0.1 in c:\users\dell\anaconda3\lib\site-packages (from pandas-profiling) (1.0.1)
Requirement already satisfied: phik>=0.11.1 in c:\users\dell\anaconda3\lib\site-packages (from pandas-profiling) (0.12.0)
Requirement already satisfied: htmlmin>=0.1.12 in c:\users\dell\anaconda3\lib\site-packages (from pandas-profiling) (0.1.12)
Requirement already satisfied: seaborn>=0.10.1 in c:\users\dell\anaconda3\lib\site-packages (from pandas-profiling) (0.11.2)
Requirement already satisfied: pydantic>=1.8.1 in c:\users\dell\anaconda3\lib\site-packages (from pandas-profiling) (1.8.2)
Requirement already satisfied: scipy>=1.4.1 in c:\users\dell\anaconda3\lib\site-packages (from pandas-profiling) (1.7.1)
Requirement already satisfied: tangled-up-in-unicode==0.1.0 in c:\users\dell\anaconda3\lib\site-packages (from pandas-profiling) (0.1.0)
Requirement already satisfied: visions[type_image_path]==0.7.4 in c:\users\dell\anaconda3\lib\site-packages (from pandas-profiling) (0.7.4)
Requirement already satisfied: pandas!=1.0.0,!=1.0.1,!=1.0.2,!=1.1.0,>=0.25.3 in c:\users\dell\anaconda3\lib\site-packages (from pandas-profiling) (1.3.4)
Requirement already satisfied: markupsafe~=2.0.1 in c:\users\dell\anaconda3\lib\site-packages (from pandas-profiling) (2.0.1)
Requirement already satisfied: multimethod>=1.4 in c:\users\dell\anaconda3\lib\site-packages (from pandas-profiling) (1.6)
Requirement already satisfied: numpy>=1.16.0 in c:\users\dell\anaconda3\lib\site-packages (from pandas-profiling) (1.20.3)
Requirement already satisfied: jinja2>=2.11.1 in c:\users\dell\anaconda3\lib\site-packages (from pandas-profiling) (2.11.3)
Requirement already satisfied: tqdm>=4.48.2 in c:\users\dell\anaconda3\lib\site-packages (from pandas-profiling) (4.62.3)
Requirement already satisfied: PyYAML>=5.0.0 in c:\users\dell\anaconda3\lib\site-packages (from pandas-profiling) (6.0)
Requirement already satisfied: missingno>=0.4.2 in c:\users\dell\anaconda3\lib\site-packages (from pandas-profiling) (0.5.0)
Requirement already satisfied: attrs>=19.3.0 in c:\users\dell\anaconda3\lib\site-packages (from visions[type_image_path]==0.7.4->pandas-profiling) (21.2.0)
Requirement already satisfied: networkx>=2.4 in c:\users\dell\anaconda3\lib\site-packages (from visions[type_image_path]==0.7.4->pandas-profiling) (2.6.3)
Requirement already satisfied: imagehash in c:\users\dell\anaconda3\lib\site-packages (from visions[type_image_path]==0.7.4->pandas-profiling) (4.2.1)
Requirement already satisfied: Pillow in c:\users\dell\anaconda3\lib\site-packages (from visions[type_image_path]==0.7.4->pandas-profiling) (8.4.0)
Requirement already satisfied: kiwisolver>=1.0.1 in c:\users\dell\anaconda3\lib\site-packages (from matplotlib>=3.2.0->pandas-profiling) (1.3.1)
Requirement already satisfied: python-dateutil>=2.7 in c:\users\dell\anaconda3\lib\site-packages (from matplotlib>=3.2.0->pandas-profiling) (2.8.2)
Requirement already satisfied: pyparsing>=2.2.1 in c:\users\dell\anaconda3\lib\site-packages (from matplotlib>=3.2.0->pandas-profiling) (3.0.4)
Requirement already satisfied: cycler>=0.10 in c:\users\dell\anaconda3\lib\site-packages (from matplotlib>=3.2.0->pandas-profiling) (0.10.0)
Note: you may need to restart the kernel to use updated packages.
Requirement already satisfied: six in c:\users\dell\anaconda3\lib\site-packages (from cycler>=0.10->matplotlib>=3.2.0->pandas-profiling) (1.16.0)
Requirement already satisfied: pytz>=2017.3 in c:\users\dell\anaconda3\lib\site-packages (from pandas!=1.0.0,!=1.0.1,!=1.0.2,!=1.1.0,>=0.25.3->pandas-profiling) (2021.3)
Requirement already satisfied: typing-extensions>=3.7.4.3 in c:\users\dell\anaconda3\lib\site-packages (from pydantic>=1.8.1->pandas-profiling) (3.10.0.2)
Requirement already satisfied: urllib3<1.27,>=1.21.1 in c:\users\dell\anaconda3\lib\site-packages (from requests>=2.24.0->pandas-profiling) (1.26.7)
Requirement already satisfied: charset-normalizer~=2.0.0 in c:\users\dell\anaconda3\lib\site-packages (from requests>=2.24.0->pandas-profiling) (2.0.4)
Requirement already satisfied: idna<4,>=2.5 in c:\users\dell\anaconda3\lib\site-packages (from requests>=2.24.0->pandas-profiling) (3.2)
Requirement already satisfied: certifi>=2017.4.17 in c:\users\dell\anaconda3\lib\site-packages (from requests>=2.24.0->pandas-profiling) (2021.10.8)
Requirement already satisfied: colorama in c:\users\dell\anaconda3\lib\site-packages (from tqdm>=4.48.2->pandas-profiling) (0.4.4)
Requirement already satisfied: PyWavelets in c:\users\dell\anaconda3\lib\site-packages (from imagehash->visions[type_image_path]==0.7.4->pandas-profiling) (1.1.1)

The description of both numerical & categorical

from pandas_profiling import ProfileReport
profile = ProfileReport(data, title="Pandas Profiling Report")

profile

print('total duplicate before',data.duplicated().sum())
print('Shape before duplicate',data.shape)
print(data.duplicated().sum())
data.drop_duplicates(inplace = True)
print('shape after droping duplicate',data.shape)

total duplicate before 240
Shape before duplicate (1599, 12)
240
shape after droping duplicate (1359, 12)

checking missing values

data.isnull().sum()

fixed acidity           0
volatile acidity        0
citric acid             0
residual sugar          0
chlorides               0
free sulfur dioxide     0
total sulfur dioxide    0
density                 0
pH                      0
sulphates               0
alcohol                 0
quality                 0
dtype: int64

• There is no null values in our data

Data Visualization

fig,ax = plt.subplots(ncols = 6,nrows = 2 ,figsize = (20,10))
index = 0
ax = ax.flatten()

for cols, value in data.items():
    if cols != 'type':
        sns.boxplot(y=cols,data = data,ax=ax[index])
        index += 1
    
plt.tight_layout(pad = 0.5,w_pad = 0.7,h_pad = 0.7)
plt.show()

fig, ax = plt.subplots(ncols = 2, nrows = 6, figsize = (20, 20))
index = 0
ax = ax.flatten()

for col, value in data.items():
    if col != 'type':
        sns.distplot(value, ax = ax[index])
        index += 1
plt.tight_layout(pad = 0.5, w_pad = 0.7, h_pad = 5.0)

• Our data seems to be easily distributed but we got some right skewed data for free sulfur dioxide,total sulphur dioxide and sulphate

Checking for the Outlier and skewnesss

from scipy.stats import zscore
z = abs(zscore(data))
print(z.shape)
data = data[(z<3).all(axis=1)]
data.shape

(1359, 12)

(1232, 12)

from scipy.stats import skew
for col in data:
    print(col)
    print(skew(data[col]))
    plt.figure()
    sns.distplot(data[col])
plt.show()

fixed acidity
0.8061619824138205
volatile acidity
0.43308445717948635
citric acid
0.2759663007760168
residual sugar
2.3718232268318102
chlorides
2.461746675680816
free sulfur dioxide
0.8662329717800913
total sulfur dioxide
1.1611815549035225
density
0.017661782151063662
pH
0.1186075802936275
sulphates
0.9494203726359354
alcohol
0.7664142127820375
quality
0.4022066765498909

• Some of the data has more skewness

data['fixed acidity'] = np.sqrt(data['fixed acidity'])
data['residual sugar'] = np.sqrt(data['residual sugar'])
data['chlorides'] = np.sqrt(data['chlorides'])
data['free sulfur dioxide'] = np.sqrt(data['free sulfur dioxide'])
data['pH'] = np.sqrt(data['pH'])
for col in data:
    print(col)
    print(skew(data[col]))

fixed acidity
0.5825645881498726
volatile acidity
0.43308445717948635
citric acid
0.2759663007760168
residual sugar
1.732402131415575
chlorides
1.461866045806564
free sulfur dioxide
0.318084567977603
total sulfur dioxide
1.1611815549035225
density
0.017661782151063662
pH
0.055277424173582446
sulphates
0.9494203726359354
alcohol
0.7664142127820375
quality
0.4022066765498909

Univariate Analysis

data['quality_map'] = data['quality'].map({3:'bad',4:"bad",5:'bad',6:'good',7:'good',8:'good'})

data['quality_map'].value_counts().plot(kind='pie',autopct='%1.1f%%')
plt.show()

• 53.6% of wine is consider as good quality
• 46.4% of wine is consider as bad quality

Bivariate analysis

plt.figure(figsize = (15,5))
sns.barplot(x = 'quality', y = 'fixed acidity', data =data)
plt.show()

• Primary fixed acids found in wine are tartaric, succinic, citric, and malic.The sum of all kind of acids determine the fixed acidity. The data spred across the quality is quit similer.

plt.figure(figsize = (15,5))
sns.barplot(x = 'quality', y = 'volatile acidity', data =data)
plt.show()

• Primary voliatile acidity is the acetic acid which have a smell and tast of vinegar is responsible for the Aroma. From the above case we can see that the less the volatile acidity is the good is wine considerd.

plt.figure(figsize = (15,5))
sns.barplot(x = 'quality', y = 'citric acid', data =data)
plt.show()

• Citric acid is used to increase acidity. It is also used to make wine taste fresh.So from the above figure we can say that higher the citric acid higher the quality of wine.

plt.figure(figsize = (15,5))
sns.barplot(x = 'quality', y = 'residual sugar', data =data)
plt.show()

• From the above figure we can say that the sweetness of all the wine is quite simiar.

plt.figure(figsize = (15,5))
sns.barplot(x = 'quality', y = 'chlorides', data =data)
plt.show()

• Cholorides refer to amount of salt present in wine. From the above figure we can say that the less the chloride the better the wine.

plt.figure(figsize = (15,5))
sns.barplot(x = 'quality', y = 'total sulfur dioxide', data =data)
plt.show()

• Sulfur dioxide is used for the preservation of wine. It helps wine in off-odors and getting brown. From the above figure we can say that the good wine has a sulphate in between 3 and 4.

plt.figure(figsize = (15,5))
sns.barplot(x = 'quality', y = 'sulphates', data =data)
plt.show()

plt.figure(figsize = (15,5))
sns.barplot(x = 'quality', y = 'alcohol', data =data)
plt.show()

plt.figure(figsize = (15,5))
sns.barplot(x = 'quality', y = 'pH', data =data)
plt.show() 

Multivariate Analysis

features_ = data.columns.values[:-1]
fig = plt.figure(figsize=(16,26))
for column , feature in enumerate(features_):
    fig.add_subplot(5,3,column+1)
    sns.boxplot(data =data, x='quality',y=feature)
plt.show()

plt.figure(figsize = (15,5))
sns.lineplot(x = 'volatile acidity', y = 'fixed acidity', hue = 'quality_map', data =data)
plt.show()

• Best quality of wine should have more content of fixed acidity and less volatile acidity

plt.figure(figsize = (15,5))
sns.lineplot(x = 'alcohol', y = 'residual sugar', hue = 'quality_map', data =data)
plt.show()

• Best quality of wine should have less content of Resudual sugar and alcohol.

sns.pairplot(data)

<seaborn.axisgrid.PairGrid at 0x1af05d1d580>

sns.pairplot(data = data,hue = 'quality')

<seaborn.axisgrid.PairGrid at 0x1af05bd3b20>

corr = data.corr()

plt.figure(figsize = (35,15))
plt.title('Correlation of all the Columns', fontsize = 20)
sns.heatmap(data.corr(), annot = True, vmin = -1, vmax = 1, center = 0, fmt = '.1g', linewidths = 1, linecolor = 'white', 
            square = True, cmap ='RdBu')

<AxesSubplot:title={'center':'Correlation of all the Columns'}>

there are no categorical variables. each feature is a number. Regression problem.

Given the set of values for features, we have to predict the quality of wine.

finding correlation of each feature with our target variable - quality

corr = data.corr()
corr['quality'].sort_values(ascending=False)

quality                 1.000000
alcohol                 0.476166
sulphates               0.251397
citric acid             0.226373
fixed acidity           0.124052
residual sugar          0.013732
free sulfur dioxide    -0.050656
pH                     -0.057731
chlorides              -0.128907
density                -0.174919
total sulfur dioxide   -0.185100
volatile acidity       -0.390558
Name: quality, dtype: float64

• The wine quality is highly correlated to the alcohol and the volatile acidity.Quality and alcohol are positively correlated.volatile acidity and quality are negatively correlated.

correlations = data.corr()['quality'].drop('quality')

Data Preprocessing

df = pd.read_csv("Wine_Quality_Predictionand_Deployment_TASK_6.csv")

def get_features(correlation_threshold):
    abs_corrs = correlations.abs()
    high_correlations = abs_corrs[abs_corrs > correlation_threshold].index.values.tolist()
    return high_correlations

taking features with correlation more than 0.05 as input x and quality as target variable y

features = get_features(0.05)
print(features)
x = df[features]
y = df['quality']

['fixed acidity', 'volatile acidity', 'citric acid', 'chlorides', 'free sulfur dioxide', 'total sulfur dioxide', 'density', 'pH', 'sulphates', 'alcohol']

x

	fixed acidity	volatile acidity	citric acid	chlorides	free sulfur dioxide	total sulfur dioxide	density	pH	sulphates	alcohol
0	7.4	0.700	0.00	0.076	11.0	34.0	0.99780	3.51	0.56	9.4
1	7.8	0.880	0.00	0.098	25.0	67.0	0.99680	3.20	0.68	9.8
2	7.8	0.760	0.04	0.092	15.0	54.0	0.99700	3.26	0.65	9.8
3	11.2	0.280	0.56	0.075	17.0	60.0	0.99800	3.16	0.58	9.8
4	7.4	0.700	0.00	0.076	11.0	34.0	0.99780	3.51	0.56	9.4
...	...	...	...	...	...	...	...	...	...	...
1594	6.2	0.600	0.08	0.090	32.0	44.0	0.99490	3.45	0.58	10.5
1595	5.9	0.550	0.10	0.062	39.0	51.0	0.99512	3.52	0.76	11.2
1596	6.3	0.510	0.13	0.076	29.0	40.0	0.99574	3.42	0.75	11.0
1597	5.9	0.645	0.12	0.075	32.0	44.0	0.99547	3.57	0.71	10.2
1598	6.0	0.310	0.47	0.067	18.0	42.0	0.99549	3.39	0.66	11.0

1599 rows × 10 columns

y

0       5
1       5
2       5
3       6
4       5
       ..
1594    5
1595    6
1596    6
1597    5
1598    6
Name: quality, Length: 1599, dtype: int64

Train and Test Split

x_train,x_test,y_train,y_test=train_test_split(x,y,random_state=3)

determining the shapes of training and testing sets

print(x_train.shape)
print(y_train.shape)
print(x_test.shape)
print(y_test.shape)

(1199, 10)
(1199,)
(400, 10)
(400,)

LinearRegression

from sklearn.model_selection import train_test_split
from sklearn.linear_model import LinearRegression

# fitting linear regression to training data
regressor = LinearRegression()
regressor.fit(x_train,y_train)

LinearRegression()

# this gives the coefficients of the 10 features selected above. 
regressor.coef_

array([ 0.01773723, -0.99256049, -0.13962865, -1.59094279,  0.00559652,
       -0.00351973,  0.76859036, -0.43741414,  0.81288805,  0.30148385])

train_pred = regressor.predict(x_train)
train_pred

array([5.33777144, 5.33826411, 5.9503318 , ..., 6.3903182 , 6.19979375,
       5.27597259])

# calculating the training and testing accuracies
print("Training accuracy :", regressor.score(x_train, y_train))
print("Testing accuracy :", regressor.score(x_test, y_test))

Training accuracy : 0.3526492871983752
Testing accuracy : 0.37860667718040975

predicting the results for the test set

test_pred = regressor.predict(x_test)
test_pred

array([5.10801475, 5.65933623, 5.90407267, 6.13461179, 5.00611866,
       5.44514691, 5.05735245, 6.15497513, 5.51919603, 5.77259374,
       5.61809366, 5.23616173, 5.23544213, 5.31968644, 6.47007277,
       5.043404  , 5.85287121, 5.19427909, 6.07727089, 6.34949018,
       6.42525555, 5.51221957, 5.8030796 , 4.93637817, 5.16618356,
       5.48255293, 5.13758624, 6.60000969, 5.88754763, 5.74133915,
       6.09716961, 6.29379754, 4.91269821, 5.88611904, 5.11007273,
       5.94574773, 6.80685536, 5.04305653, 5.25438683, 5.88611904,
       5.17406542, 4.84008442, 6.48781656, 5.40521715, 5.31105571,
       5.84484462, 5.7100681 , 5.24300809, 5.25021217, 5.46398911,
       5.08740494, 5.61369555, 6.01375792, 6.32497377, 5.47511954,
       5.36466869, 5.09234555, 4.92625623, 5.21415941, 5.08274744,
       4.79570013, 5.4377645 , 5.25237771, 5.68830391, 5.85145609,
       6.52420079, 5.38691412, 5.71775637, 5.17641417, 5.99156845,
       5.6445189 , 5.60892012, 5.74967567, 5.21702288, 5.97975854,
       5.51115845, 5.41121547, 5.6832459 , 5.63971524, 5.74133915,
       6.24163428, 5.27915822, 4.66596769, 6.04951743, 5.52401618,
       5.17823915, 5.20672986, 5.96322663, 5.50411353, 5.64866275,
       5.70105618, 5.6431575 , 5.72586828, 5.3173125 , 5.37075392,
       5.394889  , 4.82061159, 5.46006525, 5.47363879, 6.54074801,
       6.13723937, 5.61422461, 6.07821503, 6.17461539, 5.73230665,
       4.92692198, 4.73317591, 5.03851027, 5.44868797, 5.78432759,
       6.46608259, 5.47530673, 6.46876056, 5.94466642, 5.43257493,
       5.20523855, 5.34551741, 5.20749557, 6.19344578, 5.61453943,
       5.83308923, 5.20267759, 5.17702922, 5.26912156, 5.74382704,
       5.6431575 , 6.15450941, 5.89677877, 5.49186029, 5.39047629,
       5.25848318, 5.41150099, 5.70750135, 5.68001376, 6.58288921,
       5.89497164, 6.37172338, 5.72945992, 5.37936908, 5.14371952,
       5.58851063, 6.59661777, 5.24403336, 5.25594627, 5.54721935,
       5.17243958, 5.76990082, 6.10847777, 6.93985005, 4.99562031,
       5.01958735, 4.68547026, 5.82434616, 5.01708671, 5.21702288,
       5.70819215, 5.63334181, 5.33481542, 5.22220103, 5.84327644,
       5.61823023, 5.78078643, 5.51830906, 6.03898671, 5.63808482,
       5.49193476, 5.96787582, 4.82363567, 5.26331016, 5.6625652 ,
       5.73510278, 6.59570394, 5.02584187, 5.9062506 , 5.85381667,
       5.21140744, 5.68951564, 5.51649995, 5.40521715, 6.3797489 ,
       6.71642336, 4.98858413, 5.88413601, 5.75553621, 5.75508093,
       5.61677128, 5.71318209, 5.40944809, 6.05634078, 5.58276397,
       5.88366084, 6.51928901, 5.00668628, 5.4022397 , 5.18103224,
       5.17641417, 5.47511954, 5.7498414 , 5.69035164, 4.92770738,
       5.12908401, 4.98756458, 6.18242395, 5.65384546, 5.45261211,
       5.56461849, 4.99367423, 5.8451887 , 5.31537513, 5.48096741,
       5.69721582, 5.63441998, 5.69158339, 5.82596777, 5.79120742,
       6.02976291, 6.20119324, 5.2711036 , 5.04432824, 5.21865445,
       5.38560996, 4.97814926, 6.21495495, 5.44287205, 5.94855371,
       5.21554859, 6.61250215, 5.08358009, 5.29397027, 5.0345619 ,
       6.16924449, 5.78078489, 4.85807345, 5.74347603, 5.29957538,
       5.35504601, 5.17674741, 6.30056659, 5.58203663, 4.95859588,
       6.10234044, 6.03135183, 6.16235396, 5.41939397, 6.76220041,
       6.20905743, 6.08858508, 5.2295158 , 5.45339068, 5.54357835,
       5.35504601, 5.24163328, 5.74950618, 5.25054347, 6.1317778 ,
       5.42684978, 5.84929436, 4.82101128, 6.06442799, 5.06580635,
       6.43296901, 6.06017586, 5.69914195, 5.70750135, 4.90849494,
       6.00470504, 5.28204515, 5.70851728, 5.42245606, 5.12401141,
       6.4859156 , 5.3065303 , 5.97368396, 5.64309435, 6.49134579,
       6.20544121, 5.09817253, 5.47044255, 5.30164692, 5.24148223,
       6.37172338, 5.38111102, 5.41238036, 6.00040309, 4.98474017,
       5.89003832, 5.35695446, 5.18204386, 5.43418226, 5.92040958,
       4.83001281, 6.84476407, 5.17004942, 4.90738182, 5.71686848,
       5.6789345 , 5.30006437, 6.28821696, 6.8852049 , 6.5809374 ,
       5.94544069, 6.3372236 , 5.90511355, 5.56832499, 6.0037674 ,
       5.51592572, 5.47416032, 5.73230665, 5.31523101, 5.15200739,
       6.22058683, 5.30080495, 6.2234137 , 6.09823805, 5.86965547,
       5.42671619, 4.83521332, 6.04105625, 5.17954177, 5.11991572,
       6.45483786, 5.50117269, 6.75189479, 5.112668  , 5.16698378,
       5.30775351, 5.71062442, 5.10611189, 5.54132974, 5.3123985 ,
       5.16102307, 4.95778371, 5.4565424 , 5.38281735, 5.34456446,
       5.18064244, 6.11817384, 5.624629  , 5.7335454 , 6.34845774,
       5.90187012, 5.51649995, 5.69382597, 5.14872878, 5.70431206,
       6.40518994, 6.17170859, 5.46670957, 6.06382552, 5.69529022,
       6.24040985, 5.24148223, 5.5445765 , 5.04703379, 4.99563993,
       5.07398398, 5.81020153, 5.40277302, 5.9958201 , 5.1304824 ,
       6.51949038, 5.39875246, 5.49117974, 5.64759732, 5.52711217,
       5.27288466, 6.49134579, 5.83141585, 5.03473182, 5.24121201,
       5.49075176, 5.27342775, 5.5118796 , 5.04396982, 5.29167137,
       5.46007816, 5.36402691, 6.12719352, 4.99200331, 5.30164692,
       6.08655307, 5.20523855, 5.13058231, 4.66155684, 6.15823093,
       6.15917286, 6.50582017, 5.80335212, 5.74347603, 6.39828821,
       6.14436092, 5.88754763, 6.05466721, 6.03231119, 5.36683868,
       5.41989769, 5.61706715, 5.4057693 , 5.76283208, 5.2734642 ])

# rounding off the predicted values for test set
predicted_data = np.round_(test_pred)
predicted_data

array([5., 6., 6., 6., 5., 5., 5., 6., 6., 6., 6., 5., 5., 5., 6., 5., 6.,
       5., 6., 6., 6., 6., 6., 5., 5., 5., 5., 7., 6., 6., 6., 6., 5., 6.,
       5., 6., 7., 5., 5., 6., 5., 5., 6., 5., 5., 6., 6., 5., 5., 5., 5.,
       6., 6., 6., 5., 5., 5., 5., 5., 5., 5., 5., 5., 6., 6., 7., 5., 6.,
       5., 6., 6., 6., 6., 5., 6., 6., 5., 6., 6., 6., 6., 5., 5., 6., 6.,
       5., 5., 6., 6., 6., 6., 6., 6., 5., 5., 5., 5., 5., 5., 7., 6., 6.,
       6., 6., 6., 5., 5., 5., 5., 6., 6., 5., 6., 6., 5., 5., 5., 5., 6.,
       6., 6., 5., 5., 5., 6., 6., 6., 6., 5., 5., 5., 5., 6., 6., 7., 6.,
       6., 6., 5., 5., 6., 7., 5., 5., 6., 5., 6., 6., 7., 5., 5., 5., 6.,
       5., 5., 6., 6., 5., 5., 6., 6., 6., 6., 6., 6., 5., 6., 5., 5., 6.,
       6., 7., 5., 6., 6., 5., 6., 6., 5., 6., 7., 5., 6., 6., 6., 6., 6.,
       5., 6., 6., 6., 7., 5., 5., 5., 5., 5., 6., 6., 5., 5., 5., 6., 6.,
       5., 6., 5., 6., 5., 5., 6., 6., 6., 6., 6., 6., 6., 5., 5., 5., 5.,
       5., 6., 5., 6., 5., 7., 5., 5., 5., 6., 6., 5., 6., 5., 5., 5., 6.,
       6., 5., 6., 6., 6., 5., 7., 6., 6., 5., 5., 6., 5., 5., 6., 5., 6.,
       5., 6., 5., 6., 5., 6., 6., 6., 6., 5., 6., 5., 6., 5., 5., 6., 5.,
       6., 6., 6., 6., 5., 5., 5., 5., 6., 5., 5., 6., 5., 6., 5., 5., 5.,
       6., 5., 7., 5., 5., 6., 6., 5., 6., 7., 7., 6., 6., 6., 6., 6., 6.,
       5., 6., 5., 5., 6., 5., 6., 6., 6., 5., 5., 6., 5., 5., 6., 6., 7.,
       5., 5., 5., 6., 5., 6., 5., 5., 5., 5., 5., 5., 5., 6., 6., 6., 6.,
       6., 6., 6., 5., 6., 6., 6., 5., 6., 6., 6., 5., 6., 5., 5., 5., 6.,
       5., 6., 5., 7., 5., 5., 6., 6., 5., 6., 6., 5., 5., 5., 5., 6., 5.,
       5., 5., 5., 6., 5., 5., 6., 5., 5., 5., 6., 6., 7., 6., 6., 6., 6.,
       6., 6., 6., 5., 5., 6., 5., 6., 5.])

print('Mean Absolute Error:', metrics.mean_absolute_error(y_test, test_pred))
print('Mean Squared Error:', metrics.mean_squared_error(y_test, test_pred))
print('Root Mean Squared Error:', np.sqrt(metrics.mean_squared_error(y_test, test_pred)))

Mean Absolute Error: 0.48443407559851015
Mean Squared Error: 0.39380413462864783
Root Mean Squared Error: 0.6275381539226502

coeffecients = pd.DataFrame(regressor.coef_,features)
coeffecients.columns = ['Coeffecient']
coeffecients

	Coeffecient
fixed acidity	0.017737
volatile acidity	-0.992560
citric acid	-0.139629
chlorides	-1.590943
free sulfur dioxide	0.005597
total sulfur dioxide	-0.003520
density	0.768590
pH	-0.437414
sulphates	0.812888
alcohol	0.301484

• These numbers mean that holding all other features fixed, a 1 unit increase in suplhates will lead to an increase of 0.8 in Quality of wine, and similarly for the other features.
• These numbers mean that holding all other features fixed, a 1 unit increase in volatile acidity will lead to a decrease of 0.99 in Quality of wine, and similarly for the other features

Logistic Regression

from sklearn.linear_model import LogisticRegression
from sklearn.metrics import classification_report, confusion_matrix
from sklearn.model_selection import GridSearchCV, cross_val_score
from sklearn.metrics import confusion_matrix, precision_score, recall_score, auc,roc_curve,f1_score,classification_report, plot_confusion_matrix



# creating the model
model = LogisticRegression()

# feeding the training set into the model
model.fit(x_train, y_train)

# predicting the results for the test set
y_pred = model.predict(x_test)

# calculating the training and testing accuracies
print("Training accuracy :", model.score(x_train, y_train))
print("Testing accuracy :", model.score(x_test, y_test))


# Print accuracy scores
print(f'Model accuracy score: {round(accuracy_score(y_test, y_pred) * 100, 2)}%')

# classification report
print(classification_report(y_test, y_pred))

Training accuracy : 0.5829858215179317
Testing accuracy : 0.57
Model accuracy score: 57.0%
              precision    recall  f1-score   support

           3       0.00      0.00      0.00         3
           4       0.00      0.00      0.00        14
           5       0.61      0.77      0.68       166
           6       0.53      0.61      0.57       166
           7       0.00      0.00      0.00        48
           8       0.00      0.00      0.00         3

    accuracy                           0.57       400
   macro avg       0.19      0.23      0.21       400
weighted avg       0.47      0.57      0.52       400

Random Forest Classifier

from sklearn.ensemble import RandomForestClassifier
from sklearn.metrics import accuracy_score,classification_report

modelr = RandomForestClassifier()

modelr.fit(x_train,y_train)

RandomForestClassifier()

Accuracy Scorez

x_test_prediction = modelr.predict(x_test)
test_data_accuracy = accuracy_score(x_test_prediction,y_test)
print('Accuracy :',test_data_accuracy)

Accuracy : 0.72

print(classification_report(y_test,x_test_prediction))

              precision    recall  f1-score   support

           3       0.00      0.00      0.00         3
           4       0.00      0.00      0.00        14
           5       0.74      0.85      0.79       166
           6       0.71      0.73      0.72       166
           7       0.69      0.52      0.60        48
           8       0.33      0.33      0.33         3

    accuracy                           0.72       400
   macro avg       0.41      0.41      0.41       400
weighted avg       0.69      0.72      0.70       400

K-Nearest Neighbor(KNN)

from sklearn.neighbors import KNeighborsClassifier
from sklearn.metrics import accuracy_score,confusion_matrix,classification_report

accuracy = []
for i in range (1,31):
    knn = KNeighborsClassifier(n_neighbors=i)
    knn.fit(x_train,y_train)
    ypred=knn.predict(x_test)
    ac=accuracy_score(y_test,ypred)
    accuracy.append(ac)

plt.plot(range(1,31),accuracy)
plt.grid(True)
plt.show

<function matplotlib.pyplot.show(close=None, block=None)>

knn = KNeighborsClassifier(n_neighbors=15)
knn.fit(x_train,y_train)
train_pred = knn.predict(x_train)
test_pred = knn.predict(x_test)
print('Training Accuracy :',accuracy_score(y_train,train_pred))
print('Test Accuracy :',accuracy_score(y_test,test_pred))
print('Confusion Matrix : \n',confusion_matrix(y_test,test_pred))

Training Accuracy : 0.5904920767306089
Test Accuracy : 0.49
Confusion Matrix : 
 [[  0   0   1   2   0   0]
 [  0   0   6   8   0   0]
 [  0   1 109  56   0   0]
 [  0   0  83  80   3   0]
 [  0   0  14  27   7   0]
 [  0   0   2   1   0   0]]

print(classification_report(y_test,y_pred))

              precision    recall  f1-score   support

           3       0.00      0.00      0.00         3
           4       0.00      0.00      0.00        14
           5       0.59      0.44      0.51       166
           6       0.46      0.77      0.57       166
           7       1.00      0.02      0.04        48
           8       0.00      0.00      0.00         3

    accuracy                           0.50       400
   macro avg       0.34      0.20      0.19       400
weighted avg       0.56      0.50      0.45       400

• Order based on Train Accuracy-score:
• Random Forest > Logistic regression > KNN > Linear regression
• As a Result , We can see Random Forest model has the best Accuracy ration for predicting Wine Quality

• Discussion
• Behaviour of Random Forest Random Forest classifier has been the best model for wine quality prediction. Random forests use bootstrap aggregating(bagging) technique to train the model over the training set.

Deployment

import pickle
pickle.dump(modelr,open('wine_modelr.pkl','wb'))
model = pickle.load(open('wine_modelr.pkl','rb'))
result = model.score(x_test,y_test)
print(result)

0.72

Summary

• Four of the popular supervised learning models were used to predict the quality of the wine.
• We achieved a test accuracy of nearly 71% with Random forests, which was highest among all the used models.